Dallio Marcello, Romeo Mario, Di Nardo Fiammetta, Napolitano Carmine, Vaia Paolo, Olivieri Simone, Niosi Marco, Federico Alessandro
Hepatogastroenterology Division, Department of Precision Medicine, University of Campania Luigi Vanvitelli, Piazza Miraglia 2, 80138, Naples, Italy.
J Clin Exp Hepatol. 2025 Nov-Dec;15(6):102596. doi: 10.1016/j.jceh.2025.102596. Epub 2025 May 27.
BACKGROUND/AIMS: Various non-invasive tools (NITs) predicting first hepatic decompensation (HD) in advanced chronic liver disease (ACLD) enclose platelet (PLT) count. A relevant proportion of metabolic dysfunction-associated steatotic liver disease (MASLD)-ACLD patients do not show splenomegaly- and hypersplenism-related thrombocytopenia. We aimed to evaluate the performance of NITs in predicting HD according to ultrasound-assessed spleen size.
In this observational study, 148 splenic and 27 asplenic (ASP) MASLD-compensated advanced chronic liver disease (cACLD) patients were enrolled. Ultrasound artificial intelligence-based tools distinguished splenomegaly-affected patients (SAPs) and normal-spleen patients (NSPs). Albumin-Bilirubin score (ALBI) and PLT count-based NITs (PLNs) (Fbrosis-4 [FIB-4], ALBI-FIB-4, red cell distribution width-to-PLT ratio [RPR], liver stiffness measurement [LSM]-to-platelet ratio [LSM/PLTr], and ANTICIPATE ± non-alcoholic steatohepatitis [NASH]) were determined. Over 3 years, the first HD was recorded.
Limitedly to SAP, spleen area inversely correlated with PLT (relationship [R]: -0.981; < 0.0001), confirming the role of splenomegaly-related hypersplenism in conditioning thrombocytopenia. HD occurred similarly in SAPs (20.48%), NSPs (21.15%), and ASP patients (25%) (: 0.198). In NSP, PLNs showed a reduced influence on HD (FIB-4 [: 0.03], ALBI-FIB-4 [: 0.001], RPR [: 0.002], LSM/PLTr [: 0.01], and ANTICIPATE ± NASH [: 0.001]) compared to SAP. In NSP, the spleen area was inversely associated (adjusted sub-distribution hazard ratio: 0.870) and more significantly ( < 0.0001) impacted HD. Consistently, unlike SAPs, in NSPs and ASP patients, PLNs showed poor performance, and exclusively ALBI maintained a good accuracy (NSP: area under the curve [AUC]: 0.651, : 0.04; ASP patients: AUC: 0.625, : 0.03) in predicting 3-year HD.
Ultrasound-assessed spleen size affects the predictive performance of the PLNs in MASLD-cACLD patients.
背景/目的：各种预测晚期慢性肝病（ACLD）首次肝失代偿（HD）的非侵入性工具（NITs）包括血小板（PLT）计数。相当一部分代谢功能障碍相关脂肪性肝病（MASLD）-ACLD患者未表现出脾肿大和脾功能亢进相关的血小板减少。我们旨在根据超声评估的脾脏大小评估NITs预测HD的性能。
在这项观察性研究中，纳入了148例有脾的和27例无脾（ASP）的MASLD代偿期晚期慢性肝病（cACLD）患者。基于超声人工智能的工具区分了脾肿大影响的患者（SAPs）和脾脏正常的患者（NSPs）。测定了白蛋白-胆红素评分（ALBI）和基于PLT计数的NITs（PLNs）（纤维化-4 [FIB-4]、ALBI-FIB-4、红细胞分布宽度与PLT比值[RPR]、肝脏硬度测量值[LSM]与血小板比值[LSM/PLTr]以及ANTICIPATE ± 非酒精性脂肪性肝炎[NASH]）。在3年多的时间里，记录首次HD情况。
仅在SAP中，脾面积与PLT呈负相关（相关性[R]：-0.981；P<0.0001），证实了脾肿大相关的脾功能亢进在导致血小板减少中的作用。HD在SAP患者（20.48%）、NSP患者（21.15%）和ASP患者（25%）中的发生率相似（P=0.198）。在NSP中，与SAP相比，PLNs对HD的影响降低（FIB-4 [P=0.03]、ALBI-FIB-4 [P=0.001]、RPR [P=0.002]、LSM/PLTr [P=0.01]以及ANTICIPATE ± NASH [P=0.001]）。在NSP中，脾面积与HD呈负相关（调整后的亚分布风险比：0.870）且影响更显著（P<0.0001）。同样，与SAP不同，在NSP和ASP患者中，PLNs表现不佳，只有ALBI在预测3年HD时保持了良好的准确性（NSP：曲线下面积[AUC]：0.651，P=0.04；ASP患者：AUC：0.625，P=0.03）。
超声评估的脾脏大小影响PLNs在MASLD-cACLD患者中的预测性能。
